While I have been reading all of this stuff for the last 2+ years, I may have forgotten something -- but here goes anyway. This most recent discussion has been whether the crew could have done something better to identify and recover from the stall. But doesn't this beg the question of, "why did they get into the stall in the first place?"

The plane was cruising in level flight at FL350, when (presumably) the pitots iced, speed indications became unreliable, and the autopilot clicked off. The PF reacted to this "situation" by hauling back on the sidestick and zoom-climbing to FL380. This zoom-climb then caused the stall.

Now has any reason been presented explaining why the PF's reaction to the aforementioned events was logical and appropriate for an A330-trained and rated pilot? Unless such an explanation exists, I think the source of the problem (and an important contributor to why there was no successful subsequent maneuvering to a recovery) becomes fairly apparent.

Of course the initial mistake was the biggest one, but people are discussing here what could have been done if they would have realized what had happened in the first place. Normally, in an accident, it is possible to react and save the plane. But in this accident it might have been very very difficult.

Maybe he was trying to get closer to the Sun to melt the ice ?!?!

DC, it was night time so hopefully the reason he pulled up was just that he didn't know how to hand fly and decided to pull up for no good reason. Now we can prevent this from happening again by teaching people how to hand fly an airliner so it doesn't happen again.

To do this you have to hire more qualified people to fly your jets. People that could pass an ATP check ride for instance without an autopilot. Remember, that is what we did back when even with that you were lucky to get an airline job. We could even figure out how to do things in an airplane we hadn't been trained for because we were pilots, not programmers.

Only this will require paying more than 18K per year for a 300 hr pilot mill graduate. That is the part airline management hates.

Ya.....I know it was night, bubbers44. Just tryin' to make a funny.
Good luck with your proposal for new pilots !! Just glad I'm outta this racket. The highest I EVER get off the ground now is getting in and out of my 4-wheeler !!

Not sure if you can really avoid this type of freak accident by hiring more/better people. AF might have one of the more traditional approach to aviation with more and more experienced crew per aircraft than any other western airline. It's more of a mindset, less of a training issue. This will change after the final report or at least we hope so.

Even with the best training in the world (which AF probably doesn't) you can never avoid that one pilot gets it wrong. The one who gets it wrong was the one at the wrong time at the wrong place, the holes of the cheese fit together and bummm there we are. As someone correctly mentioned, there were dozends of similar incidents of pitot icing on A330/340 during the last few years and never ever did something happen.

I remember being in a simulator with a fellow captain in a well reputable TRTO when we trained TCAS/ACAS and the captain beside me turned out to be completly inproficient in his reaction, pulling up to the hard stop when he heard Mr. Airbus calling "climb, climb". When we asked him, he said he never trained TCAS before. It's as unexcusable as the AF FOs, but it happens, there is always something happening, because we are not living in a perfect world. And when several mistakes happen at the same time, people are dying. And it will happen again, but in other circumstances and with other failures. Just because there were two incapable persons in the cockpit doesn't mean that our aviation system is wrong, training is wrong and bean counters are wrong. Because 99.9999% of all pilots would have handled the incident completly sufficient and safe.

We can safely argue that there must have been a lack of knowledge and/or training in Air France, after the second pull up exercise, so this problem will be solved, hopefully, soon. Then everyone back to his/her station and then we can finally finish this discussion. Anyone agree?

Not A Freak Accident
 

A freak accident is one where an unusual situation brings it about. The situation was not unusual or unique. Different types of aircraft from various airlines transit the ITCZ multiple times each day. It just so happens that two, or maybe three, of the pilots assigned to this flight were not up to the challenge and a failure of an aircraft system occurred when these three were assigned. While it is true that the holes in the cheese did line up for this one, that happens for every accident and, in this one, there are readily identifiable contributors to the final result. I was once a passenger on a flight where the holes lined up to cause a crash but they lined up again and I was able to escape. In that case, the authorities recognized the multiple causes and took appropriate action. We must not allow anything less than full effort in correcting those multiple causes. Calling this a freak accident reduces the impetus to correct the identified problems.

No, I don't agree because there are thousands of pilots a day flying around commanding a wide body airliner with experience like these two that couldn't handle a simple loss of air speed and autosystems without crashing. Flying attitude and power must have been above their flying skills with their low time.

I would hope that we learn from this crash and make sure at least one competent pilot who can fly manually is in the cockpit. One way is to require two real captains, not type rated copilots who haven't really been given a real check ride. I know this sounds harsh but the pressure is on to pass the FO type rated pilot so they can fly over 8 hrs on international flights. They get a lot of breaks so they pass. Usually the captain check is made to guaranty he can command the aircraft.

This is my observation. It may differ elsewhere.

I see your point Dozy but I have to make exception to this... The 320 that ended in the Hudson was flown with more then basic training given by the company I am sure. Who is to say YOU will not need the "escape maneouvre" when you fly tomorrow dude?

Bank angle itself is almost irrelevant pertaining to "G"loading. I'll bet I can roll one without exceeding much more then 1.5 at the bottom. It will depend on how much and at what rate your pull (pitch). Right? (A basic acro course would sort that out rather fast)
I still say it is nice to be able to have the bank angle and pitch authority not dictated by the computers of the aircraft but a will trained crew who actually knows how to fly.
As far as the "escape maneouvre" goes it is still better to have the tools and not need them (the ability to fly beyond normal limits) then the inverse eh? However that would require more then basic flying skills and FMS management.

One thing always puzzled me is why AF447 never made any attempts to alter the flight path to navigate a major storm? All of the other 5 flights, including the one right behind (ELY10) made some attempts to avoid the storm while 447 seems to hit it head on without any corrections. Is it possible that the crew thought that they were above the storm cloud or did they choose to disregard the weather warnings?

Wait, What? ABS
 

Sorry for the OT, but, wait, Dozy, I always understood that a car (sorry, automobile) with ABS will stop in the shortest possible distance - all I have to do is stomp on the brake pedal and let the 'computer' do the rest... Is this untrue? :confused:

andianjul, you are not entirely "OT". What Dozy means is that, in modern times, we are lead to believe that "computers" do it better than you.
This is not always the case : any professional car racer will beat any ABS stopping distance any day.
The point with cars/automobile ABS is NOT achieving shorter stopping distances but to keep directional control of the car by preventing wheel locks.
To prevent wheel locks, the "computer" releases the brakes : it is easy to understand that braking action is lower when brakes are released.
On the other hand, a professional car racer is smart enough to keep braking to the maximum sustainable braking power without locking the wheels thus retaining braking power AND directional ability.

There is a big training issue here : drivers are not (enough) aware that ABS has been installed to keep directional control of the vehicle in order to avoid obstacles, not to get the best possible stopping distance.

The same is true with A/C "autobrake" : if pilots were smart enough to "feel" the maximum braking power they can get without locking any wheel, they would achieve shorter stopping distances (no brake releases).

Exactly. Well said.

"soft" limits vs "hard" limits, an old debate...

Source : Comparative study, CFIT avoidance, A332/777 (worth the reading!)

That's about statistics... if more accident because pilot overstress/stalls/overbanks the aircraft, than from aircraft preventing pilots to go to the (aerodynamical) limits, then hard protections seems better.
Other advantage of the latter are to offer constant/guaranted performance (more than pilots, in fact, who are human...)

O/T about the ABS : Braking action (from brakes to wheels) is lower when brakes are released, that's true.
But OTOH, on most surfaces, breaking is more efficient (and distance shorter) when the wheels are not locked (IIRC).
=> ABS provides:
- on all surfaces (dry, wet, mud, snow) the best directional control. 100% of the time.
- on most surfaces (dry, wet) shorter stopping distance than most of drivers can achieve. 90% of the time.
As an educated driver, I would like to be able to switch off the ABS when I drive on fresh snow. I can do that by pulling the breaker, but it's not very practical so most of the time I let it on (and know that I won't be able to achieve the shortest stopping distance).

Very good link indeed AZR.

I guess the pilot's statement says it all.

What is even more compelling, is the continuation of this recommendation:

I have repeatedly asked for such a cut-out switch on Airbus.

A little further away from CFITs, there are the two recent 0 power (or almost 0 power) accidents: Hudson and LHR.

It has been fastiduously pretended, that Sully only could succeed in putting his 320 on the hudson, due to the Airbus FBW. I believe if you can do that with a Cessna or a 320, it is just as much possible with a T7. The Aethiopian might have succeeded, if the thug was not strangling the pilot.
So no advantage to AB-FBW here.

I just wonder now, if with a 330 the LHR accident would have been as benign. Didn't the pilot oversteer the impending stall to reach the runway? In an Airbus this would have been inhibited .....

Again, to me the, although as stated "subjective", the verdict of the test pilots clearly says it all.

Airbus might simply consider to implement that cut-out switch! Their version consisting of a bulletin to tell the pilots to switch off two Prims and one Sec, is not practicable, allthough basically admitting, that such a possibility is necessary.

Who knows, it might have helped on AF447

I'll bet no manufacturer would let you try!


Where do you think the numbers in the computers came from, if not the engineering pilots (one of whom was the best in the business) who tested the thing and worked out it's limits?

Just because the protections are there doesn't mean you should go out of your way to use them - you can still be a brilliant hand-flyer and stick to your normal limits. Perhaps it would help if pilots saw the computers as there to help (which they are) rather than something to be fought against.

I'm prepared to bet that aerobatically trained pilots were incredibly rare on the line during the entire history of modern aviation - the lack thereof now has nothing to do with "FMS programming".

GF - we've had the BRB argument on the other threads - we don't need another one here. Also, the pilot of BA038 elected to stretch the glide by raising the flaps - that is all, and nothing in the Airbus flight envelope protections would have stopped him.

You can really shorten the whole Airbus-design-flaws conversation with the following observation.:

Two of Airbus' own test pilots fatally crashed an A330 one mile from where it was built after 1 minute of flight time because even they didn't understand the repercussions of ALT*, a mode less-prepared pilots will see dozens of times a day.

Any further observations?

Actually the test pilots were at the end of a long day, when most mistakes happen, and in any case - after that accident, Airbus changed the design.

.... and pilots are never at the end of a long day.

We all fly our butts off, at the back side of the clock, after 12 hours in noisy hotels or resting in crew bunks that would shock animal rights activists!
Days of legal 22 hours duty time, one inflight-rest accorded, maybe maximum 7 hours. Consecutive night shifts, paired with a superb dayshift and minimum days off, somewhat 8 a month.

And you come up with such a lame excuse.
Get real Dozy, under which stone were you hiding???? :ugh: